Negative and Complex Probability in Quantum Information


Published on

redacted by Sam

Published in: Education, Technology, Spiritual
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Negative and Complex Probability in Quantum Information

  1. 1. Negative and Complex Probability in Quantum Information Vasil Penchev1 Abstract: “Negative probability” in practice. Quantum Communication: Very small phasespace regions turn out to be thermodynamically analogical to those of superconductors.Macro-bodies or signals might exist in coherent or entangled state. Such physical objectshaving unusual properties could be the basis of quantum communication channels or evennormal physical ones … Questions and a few answers about negative probability: Why does itappear in quantum mechanics? It appears in phase-space formulated quantum mechanics;next, in quantum correlations … and for wave-particle dualism. Its meaning:- mathematically:a ratio of two measures (of sets), which are not collinear; physically: the ratio of themeasurements of two physical quantities, which are not simultaneously measurable. The maininnovation is in the mapping between phase and Hilbert space, since both are sums. Phasespace is a sum of cells, and Hilbert space is a sum of qubits. The mapping is reduced to themapping of a cell into a qubit and vice versa. Negative probability helps quantum mechanicsto be represented quasi-statistically by quasi-probabilistic distributions. Pure states of negativeprobability cannot exist, but they, where the conditions for their expression exists, decreasethe sum probability of the integrally positive regions of the distributions. They reflect theimmediate interaction (interference) of probabilities common in quantum mechanics.Key words: negative probability, quantum correlation, phase space, transformation betweenphase and Hilbert space, entanglement, Bell’s inequalities1 Institute of Philosophical Research, Bulgarian Academy of Science;;,,, http://www.philosophybulgaria.orgNegative Probability Vasil Penchev:
  2. 2. Vasil Penchev The main highlights are: 1. Negative or complex probability appears where the measure of two “parts” orof a “part” and the “whole” forms any angle. It can happen when the probability sum isgeometric in form. 2. Negative or complex probability cannot be excluded from consideration inquantum mechanics since any quantum object consists of two “parts”: wave and particle . 3. This form of probability is why the effects known as “relative rotatedmeasures” are observed in any separated quantum object as well as in the systems of quantumobjects. They reflect the immediate interaction of probabilities without any “hidden”parameters. I. Why does negative probability appear in quantum mechanics? Negative probability appears in phase-space formulated quantum mechanics;also, in quantum correlations and also in wave-particle dualism. Why? One may give a few answers to that question. The definition of negative probability expressed mathematically is the following:a ratio of two measures (of sets) which are not collinear; or physically: a ratio of themeasurement of two physical quantities, which are not simultaneously measurable. An revealing example is Heisenberg’s uncertainty measure: . Itsexpression as a negative probability would mean that the real axis of the measurement eitherof , or of has been rotated between the coupled measurements; where | | and | |: Fig. 1. The mechanism of Heisenberg’s uncertainty II | Negative and Complex Probability in Quantum Information
  3. 3. II. The “appearance” of negative probability in the phase-space formalism ofquantum mechanics can be represented schematically, using a correspondence with anordinary Hilbert-space formalism of the phase-space or by a similar correspondence betweentheir “atoms”, i.e. a phase-space cell and a Hilbert-space qubit. Fig. 2. The mapping between phase and Hilbert space reduced to the transfor-mation between a cell and a qubit The question is how the transformation between the two-dimensional cell ofphase space and the qubit of Hilbert space works: ( ) ∑ ∑ ( ) | | ( ) |∑ | , but ( ) ∑ If x and p is simultaneously measurable, then a cell can be transformed into aqubit: ⁄ ( ) ∫ *( ) ( ) ’ is the Wigner function (1932) in a contemporary view. Wigner functionin its original form is: ( ) ( )⁄ ( ) ∫ ∫ ( )* ( ) (Wigner 1932: 750). Negative and Complex Probability in Quantum Information | III
  4. 4. Vasil Penchev Wigner transformation in general: ̂ ( ) ( ) ∫ ⁄ ⟨ ⁄ | ̂| ⁄ ⟩ Reverse Wigner transformation (Weyl transformation) in general: ( ) ̂ ⟨ | ̂| ⟩ ∫ ( ) ( ) Hermann Weyl’s paper (1927) is historically first. He considered abstractly andmathematically the transformation (q) ⟺ P(x,p) , but did not interpret P(x,p) asprobability, and did not discuss the fact that it can obtain negative values. The original Weyl transformation (1927: 116-117): ( ) ∬ ( ) ( ) ∬ ( ) ( ) It follows: [] ∬∬ ( ) [( ) ( ) ] F[f] turns out to be partly analogical to Dirac’s δ-functions (Schwartzdistributions). Now, we are going to consider the time-frequency (= time-energy) reformulationof Wigner function derived by Ville (1948): “The transmission of communication signals is accomplished by means of atransmission of energy, generally of electromagnetic or of acoustic energy. … It is not energyitself which is of interest, but rather the changes in this energy in the course of time. The morecomplicated the function which represents, as a function of time, the change in voltage, cur-rent, pressure, or any other carrier, the greater is the amount of the information carried by thetransmitted energy”2 (Ville 1948: 63). When thinking about quantum information, it is especially important to look at allphysical processes as informational ones; not simply transmitting, but also being themselves2 The English translation of the citations of Ville’s paper is according to the translation by I. Selin, “Theory andApplications of the Notion of Complex Signal” – . IV | Negative and Complex Probability in Quantum Information
  5. 5. http://vasil7penchev.wordpress.comsignals. The necessary condition for such a viewpoint is the reciprocity of time and energyimplied by the Heisenberg uncertainty measure (coordinates and momentums being formallylinked to these quantities, and consequently, by the quantum/ discrete character of mechanicalmotion. We may discuss Ψ-function as a special kind of complex signal modulated by du-al processes: “Complex signals … may be considered as the result of the modulation of theirenvelope by a carrier is itself frequency modulated” (Ville 1948: 67). Very important for the interpretation of negative probability is the following cor-ollary made by Ville: “Any signal modulated by a sufficiently high frequency may be considered com-plex” (Ville 1948: 68). Consequently, any “high energetic“, physical object-like macro-body may be con-sidered as such a complex signal modulated by dual processes. A similar corollary may befound described as follows: „For any signal s(t), the function ̅( ) ( )which in general is not complex approaches the complex signal ( ) (associated with ( ) ) as increases” (Ville 1948: 68). We are inclined to consider Ψ-function in quantum mechanics “analogically” (cf.,digitally) namely as the sum of an infinite series of energetic constants. Any signal of highenough frequency, and one, which may consequently be “discussed” as complex, dual-modulated, is in fact aperiodic, enough to represent ( ). A crucial part of it happens, or maybe is better to say, is happening in a given timeinterval, which we may denote as “now and here”. Consequently, setting time-frequency analysis, on the one hand, and Ψ-function,on the other, a closer connection with the former originates from the “material particle” ofclassical physics, but a much clearer meaning within much wider limits is needed for that sce-nario: those of quantum generalization. Under quantum generalization a point-particle still remains localized, but inher-ently partitioned. By means of the “duality” it smoothly turns into a wave, and this originatesfrom an isolated and defined part of the localized point-particle, in turn affecting the totality. Negative and Complex Probability in Quantum Information | V
  6. 6. Vasil Penchev This quantum status, the whole as a set of its possibly non-additive parts, andcompletely in a Gibbsian manner, is already thought equivalent to the unity of correspondingprobable states of the whole, or “worlds”. Eventually they correlate to each other. Naturally, the question is: If the signal is a complex one, then where is the bound-ary of the two modulations; how is it to be distributed in the “space”; and by what is it is de-termined? Or in other words: Where is the boundary between the signal part that is transmit-ted by amplitude and amplitude modulation; and its alternative part that is transmitted by fre-quency, phase and frequency-phase modulation? The answer à la Ville is the following: “The proposition which we shall come to use is itself an immediate consequenceof the fact that a complex signal is characterized by the peculiarity of having a spectrumwhose amplitude is zero for negative frequencies. Now, modulating ( ) by amountsto causing the spectrum of ( ) to be translated by the amount . For a large value of , the spectrum lies entirely in the region of positive frequencies, and ( ) becomescomplex” (Ville 1948: 68-69). Using wave-particle “spectacles” reveals a “corresponding” perspective. Ville’sanswer has an obvious and simple interpretation: Amplitude and AM codes the particle prop-erties of a quantum object while frequency and phase (FM and PM) code its wave properties.That hints at Cohen’s generalization of Wigner’s function: “We now ask what the analytic signal procedure has done in terms of choosing theparticular amplitude and phase, that is, what is special about the amplitude and phase to makean analytic signal? Generally speaking, the answer is that the spectral content of the amplitude ( )is lower than the spectral content of ” (Cohen 1995: 35). So, we can restrict the spectrum of amplitude within a fixed frequency interval,corresponding to the (introduced above) “now”, by the following two steps: The first is: correspondingly linear dependence; and the second is: arbitrary time-frequency dependence: − “ ( ) is analytic if the spectrum of ( ) is contained within ( )”(Cohen 1995: 36). ( ) “ ( ) is analytic if the spectrum of ( ) is contained within ( ) and ( )the spectrum of is zero for ” (Cohen 1995: 36). The division specific for each complex signal between a part transmitted by am-plitude and a part transmitted by frequency and phase, suggests a new idea and an even more VI | Negative and Complex Probability in Quantum Information
  7. 7. http://vasil7penchev.wordpress.comdecisive generalization. Resulting from this generalization is the following interpretation: Wehave arrived at a clear division between ‘finite’ and ‘infinite’ in an arithmetical, set-theoretical and meta-mathematical sense; or between ‘syntactic’ and ‘semantic’ in a logicalview. For example, the size of amplitude, treated as a signal and thus as coded infor-mation, could be thought of as the Gödel coding of any finite syntax, or the axiomatic-deductive kernel of a certain set of tautologies, i.e. as logic: Whether in an absolute or in arelative meaning, i.e. as the “logic of A”, it can be thought of as the logic of a certain thematicdomain. A bit more detailed development of these ideas follows: After sketching the generalization of the Wigner function3, in which it is interpret-ed also as a signal and after applying a time-frequency analysis of it, we would hint at a newexposition of Ψ-function in order to allow of (and treat of) transferring results between logicand quantum information. At that time specifically, the logical equivalent of ‘negative probability’ could beinvestigated. It will be distributed along the boundary between syntax and semantics and thus,describing the transition between them. Logic understood as a formal system is depicted as “syntax” and represented by afinite set of rules for constructing formally correct sentences. However, that set is absolutelyindifferent to ‘name’ that is “semantics hidden behind symbol”. In 20th century, a usage of terms like “the logic of A” has been propagated. “A”has been understood as one or another subject domain, and consequently as semantics. Thusthe idea that the syntax depends on semantics has been involved implicitly, and vice versa: inwhich case syntax has been thought of as “logic adequate for the subject area at issue”. Now, we try to formalize how that idea acquired scientific popularity and then wewill show how the Ψ-function is isomorphic to such a formalizing. Even though, it is a new,syntactic-semantic exposition that is implied by an external, but similar approach. A general sketch of the idea could be the following: The complex coefficient is transformed and viewed as the Gödel number of thedescription; that is: as the terms of the description, which are also “coded” in terms of thecorresponding member of the basis of Hilbert space!3 “The Wigner distribution, as considered in signal analysis, was the first example of joint time-frequency distri-bution …” (Cohen 1995: 136). Negative and Complex Probability in Quantum Information | VII
  8. 8. Vasil Penchev A metaphor (which suggests also the useful idea of the relativity of ‘thing’ and‘world’) can serve us as the basis for clearing: We divide our world into “orthogonal” things,i.e. without any common intersection between any two of them, and then, describe the inves-tigated object using the set of its metaphors, that is by each one of the orthogonal things in theworld. As a humorous example, let’s say we want to make an inventory of ‘rabbit’ andthe row of things chosen by guesswork, which turns out to be suitable, is: ‘bear’, ‘house’,‘cockroach’, ‘concept’, ‘electron’, etc. The coefficient of ‘bear’ will be probably the biggest,as the scale of “resemblance” may be what we use, with ‘rabbit’ compared to ‘rabbit’ beingthe highest. Respectively, the Gödel coding of the partial description of ‘rabbit’ in the “logicof bear” will give the biggest value of the coefficient, while “likening it to”, “comparing it to”‘etc.’ more and more irrelevant things, will result in the coefficient expressing the complexityof the description tending to zero. So the ‘rabbit’ is now represented by – the most convenient form would be – asimple sum of its description in any possible world: the rabbit as a bear, the rabbit as a house,the rabbit as a cockroach, the rabbit as the thing of “notion”, the rabbit as an electron, etc. Applying such a humorous procedure, the idea of Hilbert space is represented as amathematical equivalent of a semantic network, defined by its basis, and precise quantitativecoefficients derived from the scale of the resemblance of the investigated object to anything interms of the network. The world as the set of things or of possible worlds may be discussed also by themathematical concept of category. Any one thing can be represented by any of the other things and by morphismsbetween the thing under discussion and the rest. A subspace of Hilbert space will correspond to any category of this type. If thecoding is one-to-one, then also that mapping is one-to-one. The fundamental nature of thenotion of category reveals the fundamental nature of the notion of Hilbert space. The latter isa one-to-one coded image of the former. But more interestingly: The syntactic-semantic inter-pretation of Ψ-function hints that Hilbert space may be considered as a coded image of thegeneralization of ‘category’, loosely denoted above as “semantic network”. After which thefocus is on meanings, and on the mapping morphisms between them. In reverse, if we have restricted the categories to topoi, then logic in its usualsense can be defined on them as the axiomatic-deductive “kernel of all tautologies among the VIII | Negative and Complex Probability in Quantum Information
  9. 9. http://vasil7penchev.wordpress.comset of all senses of meaning” (which are de facto pure semantic relations). At the same timelogic as a “totality”, i.e. logic as a universal and omnipresence doctrine, is also the logic of aspecific thing, implicitly taken for granted through its axioms. The deep philosophical essenceof the notion of “topos”, itself serving as the basis of ‘formal’ or ‘mathematical logic’, con-sists in allowing “topology” to be defined and hence: “discontinuity” and “continuity”. Withsuch a structure the generalization of Einstein’s principle of relativity for discrete transformsmakes sense, I mean this in the sense of quantum mechanics. Going immediately to the field of quantum information: − to apply the Skolemianrelativity of ‘discrete’ and ‘continual’ implies that the distinction between ‘continuous’ and‘discontinuous’ loses part of its sense, which leads us into stepping beyond logic, maybe tothe “language” of being: Let us divide an infinite set, e.g. that of integers, into two compact subsets, so thatany element of the initial set belongs to just one of the two subsets. The method of diagonali-zation shows that such a dividing necessarily exists, after which both subsets can be put inone-to-one mapping with the initial set (Пенчев 2009: 306). Suppose we have already con-structed a new, “actualist” representation of diagonalization, then we could build an arithme-tic version of the so-called Skolem’s paradox, as follows: A real number may be juxtaposed uniquely with any of the divided sets of theforegoing type, in such a way that when one of the sets is finite, the associated real number isa rational number, but when both are infinite, then it has an irrational one. Further, we canshow that there exists such a one-to-one mapping between the real numbers and all divisionsof that type. The set of all such divisions will be denoted by . Finally, it is evident that we can build another unique mapping between integersand the set . Since the composition of two one-one mappings is also a one-one mapping, there-fore, using an “intermediate station”, we have now built a one-one mapping between the set ofintegers and that of real numbers and therefore the former and the last are equinumerous. But, we have utilized an “actualist” version of diagonalization, which in its initial,“constructivist” version had been applied by Cantor to show that the cardinality of real num-bers is different from that of natural or rational numbers; and since the enumerable cardinalityhad been alleged as “the least” cardinality of an infinite set, it therefore followed that cardi-nality of real numbers is one of “the greater”, though not necessarily obviously the greatercardinality (that it is clearly the greater cardinality is stated by the continuum hypothesis sug-gested yet again by Cantor) (Пенчев 2009: 330-331). Negative and Complex Probability in Quantum Information | IX
  10. 10. Vasil Penchev In the “language” of being, number is not different from word, and the physical“quantity” as information is the tool by which we measure the “unity” between words andnumbers (but on the numerical, quantitative side). The natural question, which we will just put, but without deciding is: What is itscounterpart on the side of word. Is it metaphor? Let us also mention the question of whether, or to what extent, descriptions using“terms taken for granted” can be considered as the logic of the ‘name’ reserved for the set ofthose terms (which name is precisely “the class of tautologies”; which will also turn out to beincreased, since the variables are restricted to accepting values only in those terms, with re-spect to fulfilling the specified relations between those terms. Roughly speaking, ‘А is brown’will be an additional tautology, if we have limited ourselves to the logic of all brown items.) If we are able to build such a wide generalization or interpretation of the notion ofHilbert space and its element, Ψ-function, then we can generalize or interpret correspondinglyphase space. And negative probability, considered after such a generalization would also bemore widely generalized by means of generalizing the Wigner function. A milestone alongthis route is its generalization by Cohen. Such a wide generalization may be denoted as a language, signal, or informationalinterpretation. Essential to the scheme is Kripke’s conception (1975) that an exact logical notionof truth can be introduced by infinite syntax (below I will discuss whether “infinite” is neces-sary), which however remains “less” than actually infinite semantics, it seems (i.e., to be pré-cise, that in Tarski’s “as if” semantic conception of truth, “infinity” is essential, moreover aconstructive infinity, rather than semantics). After that however, it means, using Kripke’s truth, semantic instability, which ismore interesting than semantic stability (or “validity”, Lutskanov 2009: 119), that is, at least alogical tautology (law) would not be valid necessarily at the meta-level. That is to say, a newname will appear “ex nihilo” transiting from a finite to infinite syntax, or a name will disap-pear, which is the same in essence. Feferman’s notion of “reflexive closure” (Feferman 1991: 1-2) will help us toclear up which is the syntactic “kernel” shared by two possible worlds (descriptions, theories).The complement of the kernel to the set of all syntactic (analytic) statements in a given world(from the participation in forming the kernel) is semantic (= syntactic) unstable in relation toall worlds (due to the participation in forming the kernel). X | Negative and Complex Probability in Quantum Information
  11. 11. We may also mention the hypothesis that there is no ordinal between (the Fe-ferman – Schütte ordinal) and (including the case of coincidence between and ). An example may be given by intuitionism if we interpret the intuitionist theory ofinfinite set as a meta-theory towards the intuitionist theory of finite set. The law of excludedmiddle ceases to be valid. A new name (a semantic item) of “infinite set” appears. The seman-tic (= syntactic) conception of truth is presented together with a properly and only syntacticconception of truth as to infinite sets of the following kind: “Till now we have not yet knownwhether it is true, but we will ever understand”. Is such a concept of truth really syntactic?Yes, since truth is intended in the sequence of finite experiments. Being any experiment finite,it means that if we accomplish its finite procedure (algorithm), we can acquire a definite an-swer whether the tested statement (hypothesis) is true or not. Bridging our consideration to Kripke’s concept of truth, we need a new, namely“semantic-syntactic” interpretation of Ψ-function in quantum mechanics. That is not too diffi-cult as the ontology of possible logical world à la Kripke transfers the many-worlds interpre-tation of quantum mechanics into a logical language. Kripke’s concept of truth accepts theboundary between syntax and semantics as movable and it is what makes it fruitful in myopinion. Then we are going to characterize a given possible world by the proper confine-ment each to its place of syntax and semantics. Consequently, for any two possible worlds,there will exist a statement which is syntactic (“analytical”) in one of them, but semantic(“synthetic”) in the other one. The Ψ-function itself interpreted semantically-syntactically describes one and thesame thing but in different ways in any possible world, and represents a catalog of all possibledescriptions or of all expectations about its behavior (Schrödinger 1935 (49): 823-824). Reality (in the usual empirical or experimental sense of the word) is not the truecatalog, but its “change” since (or as) the catalog is invariant in anyone of all possible worlds,including also those, in which the thing is described as untrue or inexistent, through beingexactly in that boundary between syntax and semantics: a boundary being uniquely character-ised towards just that world. The semantic-syntactic interpretation of von Neumann’s theorem (1932) about theabsence of hidden parameters in quantum mechanics corresponds to “standard” quantumlogic, whose base he founded in the same book: There is nothing which is true in a finitenumber of worlds, particularly in a single world. Truth or untruth can be defined on finite setsof any world, but only on an infinite number of possible worlds. Such a standard interpreta- Negative and Complex Probability in Quantum Information | XI
  12. 12. Vasil Penchevtion is consistent also with Kripke’s truth. In addition, the boundary between (a) thing and (b)(or its) world remains absolute. A semantic-syntactic interpretation of Bell’s revision (1964, 1966), or in otherwords, defining the limits of validity for the foregoing theorem would correspond rather to“holistic semantics” (Cattaneo, Chiara, Giuntini, Paoli 2009: 193). If the interaction of possi-ble worlds is introduced consisting in the interchange of possibilities (probabilities of events)in definite rules, with respect to those between the sets of possible worlds, then the semantic-syntactic interpretation of von Neumann’s theorem is not valid. The boundary of validity isoutlined by the missing of interaction, i.e. they outline the validity only as to isolated possibleworlds. Hence, if events are progressing in more than one possible world4, then "truth" can bedefined for a finite number of or even for a single world, and the boundary between a thingand a world is not already absolute. This sketched “very wide” generalization and interpretation signposts us to Co-hen’s generalization by using its philosophical meaning drawn from the discussion of howlow and high frequencies may be managed by “division” amongst the kernels: “Therefore what the analytic procedure does, at least for signals that result in theabove forms, is to put the low frequency content in the amplitude and the high frequency con- ( )tent in the term ” (Cohen 1995: 36). Mathematically such management is accomplished by the “kernel function”: “The approach characterizes time-frequency distributions by an auxiliary function,the kernel function. The properties of a distribution are reflected by simple constraints on thekernel, and by examining the kernel one readily can ascertain the properties of distribution.This allows one to pick and choose those kernels that produce distributions with prescribed,desirable properties. This general class can be derived by the method of characteristic func-tions” (Cohen 1995: 136). Тhat kernel function may be interpreted as a filter allowing amplitude (or time)and frequency-phase component to be divided from one another. However it can also be in-terpreted as an external influence of another quantum object, i.e. representing entanglement orits degree in the deep sense of quantum information. When the generalization of the Wigner function “was subsequently realized” (Co-hen 1966: 782; 1995: 136), it was also realized that an infinite number can be readily generat-ed from:4 It means: The change of probability or the progressing of an event in one of the worlds changes the probabilityof progressing an event in another possible world. XII | Negative and Complex Probability in Quantum Information
  13. 13. ( ) ∭ ( ) *( ) ( )where ( ) is an arbitrary function called the kernel 5 by Claasen and Mecklenbrauker6”(Cohen 1989: 943). The meaning can be made crystal clear and is complemented by the next twoschemata: Fig. 3. The relationship between the Wigner distribution function, the auto- correlation function and the ambiguity function (from Wikipedia: Cohens class distribution function ) The notations are the following: ( ) ∫ ( ) ( ) ( ⁄ ) ( ⁄ ) ( ) ∫ ( ⁄ ) ( )5 “In general the kernel may depend explicitly on time and frequency and in addition may also be a functional ofthe signal” (Cohen 1989: 943, footnote 3).6 T. A. C. M. Claasen and W. F. G. Mecklenbrauker, "The Wigner distribution – a tool for time-frequency signalanalysis; part III: relations with other time-frequency signal transformations,” Philips J. Res., vol. 35, pp. 372-389, 1980. Negative and Complex Probability in Quantum Information | XIII
  14. 14. Vasil Penchev The first one is Wigner function transformed into parameters of time and frequen-cy from the parameters of coordinates and momentums. The second one is the autocorrelationfunction, and the third one the ambiguity function. t is time, and the time of another instant (e.g. the function „decelerated” or „ac-celerated”). f is frequency in relation to the instant t, and η frequency in relation to the instantof . * is the operation of convolution. Obviously, the autocorrelation function expresses immediately interaction be-tween different time instants, and the Wigner function and ambiguity function through theagency of frequency, do the same. Hence, it is clear that the introduction of negative probabil-ity for some small regions of phase space in the usual representation of the Wigner functionreflects indirectly the interaction of different time instants, which is “unpacked” in consecu-tive order by transformation of the parameters to time-frequency, and ultimately, by the inter-actions connection with the autocorrelation function. Of course, classical physics takes for granted that such a type of interaction is ab-sent, for its availability would produce retro-causality. The additional kernel function of Co-hen’s generalization allows for the degree of interaction between the instants to be regulated,particularly to be concentrated onto some of them. The Wigner function is the special casewhen the kernel function being 1 does not exert an effect − but it does not correspond, as wehave seen, to the absence of negative probability! The method and degree of separation between the time instants is regulated by thefundamental constant of light velocity in vacuum. From such a viewpoint Minkowski spacerepresents the area of autocorrelation, i.e. of possible physical interaction. It is described usingduality in two alternative ways: without correlations, i.e. by diffeomorphisms according toEinstein’s principle of relativity; or with correlations, i.e. in the standard way of quantum me-chanics, according to which the Cohen function is reduced to the Wigner one since the kernelfunction is 1. The next figure displays how kernel function can be used as a filter for the sepa-ration, with respect to increase, either of the interaction or the degree of distinction betweenthe instants: XIV | Negative and Complex Probability in Quantum Information
  15. 15. Fig. 4. What is the benefit of the additional kernel function? The figure shows the distribution of the auto-term and the cross-term of a multi-component signal in both the ambiguity and the Wigner distribution function. (Wikipedia: Cohens class distribution function ) Furthermore it is obvious that if “straight” Wigner function can be generalized,then analogically the reverse transform of Weyl can be, and that was done by Cohen in a laterpaper: Cohen’s generalization of the Weyl transform is the following: “ ( ) ∬ ̂( ) ( ) ,where ( ) is a two dimensional function called the kernel 7. The kernel characterizes aspecific transform and its properties” (Cohen 2008: 260). It is well to keep in mind the manner of Cohen’s thought while discussing Groe-newold’s statistical ideas (1946). Their essence was recapitulated by Groenewold himself asfollows: “Our problems are about:  α the correspondence a⟷a between physical quantities a and quantum operators a (quantization) and  β the possibility of understanding the statistical character of quantum mechan- ics by averaging over uniquely determined processes as in statistical classical mechanics (interpretation)” (Groenewold 1946: 405). α, the correspondence a⟷a (quantization), in fact, generates two kinds of prob-lems about the physical quantities a:7 Cohen 1966. Negative and Complex Probability in Quantum Information | XV
  16. 16. Vasil Penchev  a is not a continuous function (it is either a continuous one or a generalized one, i.e. a distribution);  There exists quantities ai whose product is not commutative. The difficulties in α (the quantization of physical quantities) reflect in the samefashion in β (statistical description). The negative probability of some states does appear, butthey are easily interpreted physically by the regions of partial overlap between orthogonalprobabilistic distributions. The main ideas of Moyal’s statistical approach (1949) would be represented by afew significant textual citations from his works: “Classical statistical mechanics is, however, only a special case in the general the-ory of dynamical statistical (stochastic) processes. In the general case, there is the possibilityof ‘diffusion’ of the probability ‘fluid’, so that the transformation with time of the probabilitydistribution need not be deterministic in the classical sense. In this paper, we shallattempt to interpret quantum mechanics as a form of such a general statistical dynamics”(Moyal 1949: 99). Fig. 5. Moyal’s statistical approach According to Moyal, “phase-space distributions are not unique for a given state,but depend on the variables one is going to measure. In Heisenbergs words8 , the statisticalpredictions of quantum theory are thus significant only when combined with experiments8 “Heisenberg, W. The physical principles of the quantum theory (Cambridge, 1930), p.34.” XVI | Negative and Complex Probability in Quantum Information
  17. 17. http://vasil7penchev.wordpress.comwhich are actually capable of observing the phenomena treated by the statistics’” (Moyal1949: 100); Fig. 6. Moyal’s statistical approach (1949) Statistical Description description by Ψ-functions Non-standard Boltz- Gibbs mann ensemble ensemble Fig. 7. Statistical or Ψ-function description We should emphasize the significance of ‘spin’, which is a characteristic physicalquantity of any quantum object (in contrast to any object of classical physics), for makingapparent and clear „probability dependence”: „symmetry (or antisymmetry) conditions intro- Negative and Complex Probability in Quantum Information | XVII
  18. 18. Vasil Penchevduce a probability dependence between any two particles in B. E . (or F. D.) ensembles evenin the absence of any energy interaction. … It is this dependence which gives rise to theexchange energy between the particles when they interact” (Moyal 1949: 116): ( ) ̅̅̅̅̅̅ ̅ ̅ ∑ | | ( ) ̅̅̅̅̅̅ ̅ ̅ ∑ | | ∬ ( ) ∬ ( ) (Moyal 1949: 116, eq. 14.8) The parameter γ accepts the following values: in a Maxwell – Boltzmann ensem-ble (the classical case) γ=0; in a Bose – Einstein ensemble: γ=1; in a Fermi – Dirac:γ= –1. ni, nk are average frequencies of the number of articles ai, with respect to ak, occupyinga given micro-state αi , with respect to αk (Moyal 1949: 114). Speaking of the three basic abstract mathematical spaces of our physics, namelyHilbert, phase, and Minkowski (or pseudo-Riemannian) space, they form a eureka “triangle”prompting the notion that one of its side is missing: Fig. 8. Abstract mathematical spaces and transformations between them XVIII | Negative and Complex Probability in Quantum Information
  19. 19. III. Negative probability in quantum correlations: In the light of quantum in-formation and quantum correlations the studied by it, the battle for or against “hidden parame-ters” in quantum mechanics can be interpreted and reformulated as local “hidden parameters”(causality) against nonlocal ones (quantum correlations). The initial notions should be sought for in: the famous „paradox” (or argument, infact and essence) of Einstein, Podolsky, and Rosen and Schrödinger’s “cat paradox paper”, in1935. The quantum correlation deducted and discussed in those articles implies negativeprobability in the final analysis. Fig. 9. Gedanken experiment Einstein – Podolsky – Rosen (1935)9 Negative probability appears “effectively”, i.e. by the restriction of the degrees offreedom (DOF) of any correlating quantum object by the others. The mechanism of suchtransformation is bellow discussed. The quantum allegory of Schrödinger’s alive-and-dead cat helps us to understandthe actual state by restricting DOF of any possible states, and consequently, working back-9 The second line of the figure translated in English: Leads to two different states of the system II ← an alterna-tive choice between the measuring of: (see the arrows towards either bellow or above); in details, Penchev 2009:55ff). Negative and Complex Probability in Quantum Information | XIX
  20. 20. Vasil Penchevwards, the quantum correlations to be suggested between the states forming any quantum su-perposition; and hence, presents a possible form of the mechanism of the mystic “collapse” ofwave packet during the real process of measurement. Fig. 10. Schrödinger’s (1935: (48) 812) poor “cat” Von Neumann’s theorem (1932) about the absence of hidden parameters in quan-tum mechanics underlies both quantum correlation and quantum superposition. Its conclusionis: “There are no ensembles which are free from dispersion. There are homogeneous ensem-bles…” (Neumann 1932: 170)10. Consequently, there are no homogeneous ensembles by con-tradiction, i.e. for example those of a single quantity being free of dispersion. Any quantityhas dispersion, which is not due to any cause, to any hidden variable. The premises of thetheorem explicated by von Neumann himself are six (A, B, α, β, I, II) as follows: A: ������≥0 ⇒ Erw(������) ≥ 0 (Neumann 1932: 165); B: Erw(a. ������+b.������+…) =a.Erw(������)+ b.Erw(������)+ …, where a, b (Neumann 1932: 165); α: ������ is a dispersion freequantity ≝ Erw(R1)=1; [Erw(������, φ)=(R φ, φ)] (Neumann 1932: 166); β: ������ is a homogenousone≝{ a, b , a+b=1, Erw(������)= aErw(������)+bErw} ⇒ {a =0 ⤄ b =0} (Neumann 1932: 166);I. {������⟼R} ⇒{f(������) ⟼f(R)} (Neumann 1932: 167); II. {������ ⟼R, ������ ⟼ S, …} ⇒{������ + ������+…⟼R+S+…} (Neumann 1932: 167). We should emphasize the correspondence of “one-to-one” between a mathemati-cal entity as hypermaximal Hermitian operator and a physical entity as quantity. “There cor-responds to each physical quantity of a quantum mechanical system, a unique hypermaximal10 Here and bellow the translation of the textual citations from von Neumann’s book from German to English isaccording to: J. von Neumann. 1955. Mathematical Foundations of Quantum Mechanics. Princeton: UniversityPress. XX | Negative and Complex Probability in Quantum Information
  21. 21. http://vasil7penchev.wordpress.comHermitian operator, as we know … and it is convenient to assume that this correspondence isone-to-one – that is, that actually each hypermaximal operator corresponds to a physical quan-tity.” (Neumann 1932: 167) What is the connection between von Neumann’s theorem and negative probabili-ties? By introducing negative probability, expectation is not an additive one in general, thepremises of the theorem are not fulfilled, and the deduction is not valid. Fig. 11. Von Neumann’s theorem and negative probability Here are a few equivalent expressions for the boundary of the validity of vonNeumann’s theorem: 1. Non-negative probability 2. Orthogonal possible states 3. Separated “worlds” 4. An isolated quantum system 5. The additivity of expectation Bell’s criticism (1966) about von Neumann’s theorem partly rediscovered GreteHermann’s objections (1935) and is very important for revealing the connection between cau-sality, quantum correlation, and negative probability: “The demonstrations of von Neumannand others, that quantum mechanics does not permit a hidden variable interpretation, are re-considered. It is shown that their essential axioms are unreasonable. It is urged that in furtherexamination of this problem an interesting axiom would be that mutually distant systems areindependent of one another” (Bell 1966: 447). “His essential assumption is: Any real linearcombination of any two Hermitian operators represents an observable, and the same linearcombination of expectation values is the expectation value of the combination. This is true forquantum mechanical states; it is required by von Neumann of the hypothetical dispersion freestates also” (Bell 1966: 448-449). Negative and Complex Probability in Quantum Information | XXI
  22. 22. Vasil Penchev The idea of Bell’s inequalities (1964) can be expressed by negative probability.Since von Neumann’s theorem is valid only about nonnegative probability (expectation addi-tivity), and quantum mechanics permits negative probability, the idea is that the domain of thetheorem validity is to be described by an inequality of the expectation of two quantities (thespin of two particles) according to the EPR conditions: (⃗ ) | ( ⃗) ( )| (Bell 1964: 198). The last inequality can be rewritten by means of an alleged hidden parameter λ, bywhich the connection with von Neumann’s theorem about the absence of any hidden parame-ters such as λ becomes clearer: 1 + P[b(λ),c(λ)] ≥ |P[a(λ),b(λ)] − P[a(λ),c(λ)]| In the light of von Neumann’s theorem, Bell’s paper (1966) reveals that an even-tual violation of the inequality above would require generalizing the notion of hidden parame-ter introducing nonlocal one. The main steps are: 1. Von Neumann’s theorem as well as the theories of hidden parameters interpretsthem as local ones implicitly. 2. Bell’s inequalities discuss the distinction between local and nonlocal parameterbecause quantum mechanics allows nonlocal variables. Fig. 12. Non-local hidden parameter XXII | Negative and Complex Probability in Quantum Information
  23. 23. The notion of non-local hidden parameter suggests the notion of the externality ofa system considering it as a nonstandard, namely ambient, or “environmental” part of the sys-tem. Not any external neighborhood, but only a small one (of the order of a few ℏ), at thatcorrespondingly, only in phase space. In the corresponding small neighborhood in Minkowskispace, Lorentz invariance is not valid. The following three figures show that: Fig. 13. Groove uncertainty! Fig. 14. A small neighborhood in Minkowski space Negative and Complex Probability in Quantum Information | XXIII
  24. 24. Vasil Penchev Fig. 15. Lorentz invariance and uncertainty relation in a small neighborhood inMinkowski space The topic may be illustrated by the notion of an absolutely immovable body: Hei-senberg’s uncertainty excludes any absolutely immovable body as well as any exactly con-stant phase volume. A body is outlined rather by an undetermined “aura” or “halo” than by asharp outline. The aura is outlined within phase space and its magnitude is comparable withthe Planck constant. It consists of the states of negative probability, which “push out” thestates of any other bodies beyond the outline region: Fig. 16. The “halo” of negative probability in phase space XXIV | Negative and Complex Probability in Quantum Information
  25. 25. The next table shows the appearance of negative probability when quantum me-chanics is reformulated from the standard language of Ψ-functions into that of a standard sta-tistical description containing, however, nonstandard, or “external to” the whole, „parts” of it: Fig. 17. Again about the comparison of a Gibbs and of a non-standard Boltzmannensemble Such a kind of explanation can be applied also to true Bell’s inequalities: Fig. 18. A mechanism of violating Bell’s inequalities (Bell 1964) Negative probability cannot help but violate Bell’s inequalities: Negative and Complex Probability in Quantum Information | XXV
  26. 26. Vasil Penchev Fig. 19. How does negative probability violate Bell’s inequalities? The notion of effective probability aids us in bridging physical ‘interaction’ andmathematical ‘probability’. It is effective probability that exerts impact from a givenprobability onto another. Effective probability is a probabilistic “force” by means of whichboth the probabilities interact (or more than 2). Since Dirac (1942: 8), negative probability hasbeen thought of just as an effective probability rather than a “real” probability of whatever:Negative probability has been likened to money since it assists “the balance”, while beingnon-existent at that! Following after Mermin (1998), we could at least question the role ofnegative probabilities or of any description of probability as a new, nonstandard, but maybeomnipresent, universal and all-embracing substance. This substance includes matter and ener-gy, which are ordinary and generally accepted substances in physics. Both the aspects symbolized by the EPR argument and Schrödinger’s cat, respec-tively of two or more “probabilistically” interacting systems and of the superposition of thestates of a single and isolated system suggest a Skolemian type of relativity between isolationand interaction as well as an extension of von Neumann’s theorem. The corollary, after gener-alizing from an isolated to two or more interacting systems is the immediate interaction ofprobabilities: XXVI | Negative and Complex Probability in Quantum Information
  27. 27. Fig. 20. The notion of effective probability (immediate probabilistic interaction) The necessary and sufficient condition of immediate probabilistic interaction is:shared common possible states of nonzero probability. The same equivalence refers to thesuperposition of all the states of an isolated system. The notion of “effective probability” orthe more particular, but correlate one of “negative probability”, are the basis of the perceptiblecommonality, that “sameness” of the Skolemian relativity: Fig. 21. How do probabilities interact? Fig. 22. The notion of effective probability (immediate probabilistic interaction) Negative and Complex Probability in Quantum Information | XXVII
  28. 28. Vasil Penchev However the halo of negative probability “pushes away”: − the system itself; anyof the 2; and the states of positive probability, too. They become relatively more probable.Consequently, a Skolemian or an Einsteinian relativity, by means of negative probability, re-veals the intimate mechanism of any physical interaction manifesting itself. In the final analy-sis by restricting DOF of any system participating in the interaction in question any physicalinteraction manifests itself. We have two kinds of description, which are equivalent: Fig. 23. A comparison of the statistical and standard formalism by means of “effective probability” Negative probability is only one of two ways to represent physical reality. Whichway corresponds more to the Boltzmann statistical consideration than to the Gibbs one? Foreach, there exist two kinds of ontological projection: XXVIII | Negative and Complex Probability in Quantum Information
  29. 29. Fig. 24. The statistical vs. standard formalism The statistical formalism enables the simultaneously immeasurable quantities ofthe standard formalism to be calculated together. Kochen-Specker’s theorem shows that ho-mogenous quantities have dispersion, even in quantities simultaneously measurable in amathematical sense, and correspondingly, that there cannot be “hidden parameters”. Also, Kochen – Specker’s theorem can be discussed as a generalization of vonNeumann’s theorem in the following way:  Von Neumann’s theorem (1932: 157-173) covers isolated systems and simul-taneously immeasurable quantities.  Bell’s theorem (or inequalities, 1964) clarifies the effect and interpretation ofthe absence of hidden parameters in interacting systems.  Kochen – Specker’s theorem deals with isolated systems and simultaneouslymeasurable quantities as follows: the dispersion of homogenous quantities is conditioned byquantum “leaps”. The two authors formalized the notion of simultaneous measurability: Kochen andSpecker (1967: 63 …) interpreted simultaneous measurability as the availability of a commonmeasure which required the measure of the set of points of discontinuity (quantum leaps) tobe zero, i.e. simultaneously measurable quantities should not be continuous, but almost con- Negative and Complex Probability in Quantum Information | XXIX
  30. 30. Vasil Penchevtinuous. They proved on such a condition that homogenous quantities have necessarily a dis-persion, i.e. there are no hidden parameters. Next, there isn’t any homomorphism of the algebra of statements about commut-ing quantum quantities into Boolean algebra. An immediate corollary follows: there is no mapping of a qubit even of two sim-ultaneously measurable quantities, into a bit. There is a propositional formula which is a classical tautology, but which is “nottrue” after substituting quantum propositions. Kochen – Specker’s theorem is connected to the Skolemian relativity of the dis- 2crete and continuous: In § 5 of their paper a model of hidden parameter in of the particleof spin ½ is constructed; however no such model exists according von Neumann’s theorem(Kochen, Specker 1967: 74-75). That model of hidden parameter is isomorphic, in fact, to themapping of a qubit into a bit. The immediate corollary comes to bear: There is no mapping of a qubit of twosimultaneously measurable quantities into a bit. Thus an explanation would be that the notionof simultaneous measurability introduces implicitly Skolemian relativity. In fact, wave-particle dualism in quantum mechanics is a form of the Skolemianrelativity of the discrete (in quantum mechanics) and continuous (in classical physics). That’swhy a Skolemian type of relativity appears also between the availability and absence of hid-den parameters. A qubit can and cannot (à la Skolem) be represented as a bit (Kochen, Speck-er 1967: 70, esp. “Remark”). IV. Negative probability for wave-particle dualism We should return to Einstein’s paper (1905I) about mass and energy: “Gibt ein Körper die Energie L in Form von Strahlung ab, so verkleinert sich sei-ne Masse um L/V2. Hierbei ist es offenbar unwesentlich, daß die dem Körper entzogeneEnergie gerade in Energie der Strahlung übergeht, so daß wir zu der allgemeineren Folgerunggeführt werden: Die Masse eines Körpers ist ein Maß für dessen Energieinhalt; ändert sich dieEnergie um L, so ändert sich die Masse in demselben Sinne um L/9.1020, wenn die Energie inErg und die Masse in Grammen gemessen wird” (Einstein 1905I: 641). (V=3.1010 [cm/s] isthe speed of electromagnetic radiation in vacuum.) Next, we are going to juxtapose the quote above with a no less famous paper ofthe same year (Einstein: 1905Ü) about quanta and energy, by means of which I search for theconnection between mass and quanta: XXX | Negative and Complex Probability in Quantum Information
  31. 31. In contemporary designations its content may be abstracted by the formula: Е=ℏν.In original designations and Einstein’s words: “monochromatise Strahlung … wie ein diskon-tinuerliches Medium verhält, welches aus Energiequanten von der Gröse Rβν/N besteht”(Einstein 1905Ü: 143); “das erzeugte Licht aus Energiequanten von der Gröse (R/N)βν be-stehe, vobei ν die betreffende Frequenz bedeutet” (Einstein 1905Ü: 144) „R die absoluteGaskonstante, N die Anzahl der „wirklichen Moleküle" in einem Grammäquivalent … be-deutet” (Einstein 1905Ü: 134), and “β= 4.866.10-11” (Einstein 1905Ü: 136). „Monochroma-tische Strahlung … verhält sich …, wie wenn sie sus voneinander unabhängungen Ener-giequanten von der Gröse Rβν/N bestünde” (Einstein 1905Ü: 143). Light is discrete (corpuscular) in its interaction with matter, but it is continuous(wave) as a medium „by itself” which is propagated in space. Quantum mechanics transferredinitially that property formulated about electromagnetic radiation to all the quanta, and in fact,to all the physical objects. Afterwards the separation of different relations has been abandonedreplacing it by the Ψ-function description. The hypothesis of hidden parameters from such aviewpoint, in fact, conserves the original and already clearly formulated opinion of Einstein(1905) that the two contradicting aspects, continuity and discreteness, are to be separated, andthus to be separated into different relations. Consequently, energy à la Einstein is already determined in two incompatibleways: as a continual quantity of mass and as a discrete number of quanta. Mechanical energyin classical physics is the sum of kinetic and potential energy: E=Ek(p)+Ep(x). Transferredinto quantum mechanics, it is the sum of momentum and location functions, which are simul-taneously non-measureable. That sum is measured by a third way, by frequency (Neumann1932: 256, Anm. 164). Von Neumann (1932: 164) had already pointed out that non-commutability refers only to the multiplication, not to the addition of operators and had giventhe example (just above) of energy as a sum of simultaneously immeasurable quantities. Apossible solution is to define the presence of discreteness (quanta) in quantum mechanics tobe ascribed to the sum or availability of non-commuting (simultaneously immeasurable)quantities. Before passing to Einstein’s principle of general relativity, let us attempt an order-ing of the “mess” of notions: Negative and Complex Probability in Quantum Information | XXXI
  32. 32. Vasil Penchev Noncommuting q-tities Relativity: Lorentz The violation of Bell’s inequalities invariance Interacting quantum systems Noncommesurable q. Commuting q-tities Neumann’s theorem Commeasurable q. Wave-corpuscular dualism; Classical Non-local causality the absence of hidden para- continuous meters; the KS theorem; isolated quantum systems quantities The absence Local of causality causality Fulfilling Bell’s inequalities Negative probability Energy Information as a physical quantity Fig. 25. An attempt at ordering the notions Einstein (1918) formulated the first two principles of general relativity as follows:„a) Relativitätsprinzip: Die Naturgesetze sind nur Aussagen über zeiträumliche Koinzidenzen;sie finden deshalb ihren einzig natürlichen Ausdruck in allgemein kovarianten Gleichungen.b) Äquivalenzprinzip: Tragheit und Schwere sind wesensgleich” (Einstein 1918: 241). Einstein reformulated the principle of relativity in mathematical language as theinvariance of physical laws towards diffeomorfisms. In this way discrete transformations wereexcluded. However they are the essential subject of quantum mechanics. Correspondingly, theuniting of quantum mechanics and relativity may be researched in terms of a generalization ofinvariance to a wider class of morphisms, which should include discrete ones. The wave-particle dualism considered as a definite type of generalization of theprinciple of relativity introduces negative probability. The following hypothesis can be ad-vanced: that any relevant generalization of the principle of relativity which includes discretemorphisms should introduce negative probability explicitly or implicitly. The principles of general relativity (1918): The original variant (Einstein 1918: 243) of the basic equation is: ( ). The repaired variant (Einstein 1918: 243) with the cosmological constant is: ( ). XXXII | Negative and Complex Probability in Quantum Information
  33. 33. Here follows a possible generalization of the principle of relativity and negativeprobability. Its three sequential levels are as follows: 1. ( ) 2. ( ) 3. ( ) There exists a link between any possible generalization of relativity and a rejec-tion (or a generalization) of so-called (by Einstein) Mach’s principle: „c) Machsches Prinzip: Das G-Feld ist restlos durch die Massen der Körper be-stimmt. Da Masse und Energie nach den Ergebnissen der speziellen Relativitätstheorie dasGleiche sind und die Energie formal durch den symmetrischen Energietensor (Tμν), beschrie-ben wird, so besagt dies, daß das durch den Energietensor der Materie bedingt und bestimmtsei” (Einstein 1918: 241-242) It seems that the colligation of general relativity (introducing negative probabili-ties) implies restricting the validity or generalization of Mach’s principle: The negative prob-ability restriction of the degrees of freedom of a part of a system can be equivalently equatedwith energy and hence, with mass. For the formula Е=ℏν the frequency ν should interpret asthe density of informational exchange between the parts of the system (i.e. bits per time). Non-commutability, wave-particle dualism (and by it, the rejecting Mach’s prin-ciple material or physical information), and “hidden parameters” can be juxtaposed straight-forwardly:  Non-commutability is a sufficient, but not necessary condition of discreteness(quanta) and hence, of wave-particle dualism.  Wave-particle dualism is a necessary and sufficient condition of the absence oflocal hidden parameters as well as of non-local ones.  An introduction of non-local hidden parameters à la Bell implies the violationof wave-particle dualism: either only “waves”, or only “particles”.  Mathematical non-commutability is interpreted as simultaneous non-commeasurability of corresponding quantities.  The commeasurability of physical quantities is interpreted as mathematicalcommeasurability, i.e. as the availability of general measure.  Being interpreted, the availability of common measure covers the cases both ofcontinuous quantities and of their discrete leaps. Negative and Complex Probability in Quantum Information | XXXIII
  34. 34. Vasil Penchev Returning now to the fig. 25 and our attempt at ordering the notions, we can al-ready emphasize Kochen-Specker’s theorem as a generalization of von Neumann’s theorem: Fig. 26. Von Neumann’s and Kochen – Specker’s theorem A Skolemian relativity of “hidden parameters” is clearly seen: If the KS theoremis equivalent to wave-particle dualism, then the very absence of hidden parameters in quan-tum mechanics should consider as equivalent with a Skolemian relativity of the continual anddiscrete. However their counterexample concerning also their proper construction shows thateven the true absence of hidden parameters is relative. Consequently, even the true relativityof the continual and discrete is also relative, which represents “super-Skolemian” relativity. We may introduce the term of “contramotion” borrowed from A. and B. Strugat-sky’s novel “Monday begins on Saturday” (1965) to illustrate the opposite directions of thediscrete and the continual of wave-particle dualism: XXXIV | Negative and Complex Probability in Quantum Information
  35. 35. Fig. 27. Janus Poluektovich and his parrot named Photon The novel represents a two-faced character, unambiguously called Janus. Eachface however is then separated into isolated bodies which are never together in the same placeand time. Each body is a different man, respectively Janus-Administrator (A-Janus) andJanus-Scientist (S-Janus). In a distant future Janus will manage (A-Janus’s point of view) orhas managed (S-Janus’s) to reverse himself towards the arrow of time. On his shoulder, hisparrot Photon will stay (or has stayed) at that instant, that’s why it will transform (or hastransformed) in a “contramotion”, too. Negative and Complex Probability in Quantum Information | XXXV
  36. 36. Vasil Penchev Fig. 28. Contramotion: Janus Poluektovich and his parrot named Photon against the arrow of time So A-Janus with all the rest is moved “correctly” in time arrow, but S-Janus andPhoton being contramotioners are moved “inversely” towards the arrow of time. Howeverthere is an over ruling circumstance: Contramotion is discontinous. Exactly at midnightS-Janus takes his parrot with him, stays alone in a deep rest, and … both jump into the daybefore instead of passing in the next day as all normal people, incl. A-Janus. Consequently, the personage of Janus with his two faces embodied in two personsmay be interpreted as an allegory of wave-corpuscular dualism pointing out the possibility ofthe two aspects being disjunctively separated forwards or backs in time: XXXVI | Negative and Complex Probability in Quantum Information
  37. 37. Fig. 29. Wave-corpuscular dualism, contramotion and negative probability Discontinuous contramotion implies negative probability since the two measures,correspondingly of the discrete and of the continuous motion, are directed oppositely. This way, we can create and suggest a new, quantum, fable, “Parrot and Cat”,combining Strugatsky’s parrot and Scrödinger’s cat. The condition of its precept or moral is:Contramotion is the sufficient, but not necessary condition of coherent superposition. Ourallegory would be: Schrödinger’s cat has eaten Photon, the parrot – contramotioner (ratherblack humor). P. Dicac’s conception (1942) helps us to put the question about the ontologicalstatus of negative probability, or figuratively said, whether negative probability might be“eaten”, and in such a way, “transferred” being rather a property of a material thing (like theparrot named “Photon”) than a relation between two or more things: “Negative energies and probabilities should not be considered as nonsense. Theyare well-defined concepts mathematically, like a negative sum of money, since the equationswhich express the important properties of energies and probabilities can still be used whenthey are negative. Thus negative energies and probabilities should be considered simply asthings which do not appear in experimental results” (Dirac 1942: 8) As to bosons “there is theadded difficulty that states of negative energy occur with a negative probability” (Dirac 1942:1). Obviously, Dirac’s position is categorically in favor of only the relative nature of negativeprobability, or in his words, only “like a negative sum of money” in its balance. I would mention Feynman’s article (1991), which is often cited. He did not go be-yond Dirac’s approach of introducing negative probabilities only conventionally, in the course Negative and Complex Probability in Quantum Information | XXXVII
  38. 38. Vasil Penchevof calculations, in an analogy of “negative money”. He gave many examples from classicaland quantum physics. Negative probability in them meant that the happening of an event de-creased of the realizing of another: i.e. negative probability could have only relative character. Much more interesting is Pauli’s (2000: 71-72 ) consideration about negativeprobability: On the subject of the theory of Gupta-Bleurer Pauli discussed the formalism of“negative probabilities”, which they had introduced. The norm and the expectation had beengeneralized correspondingly as follows: ∑ ∑ ; ∑ ∑ 〈 〉 ∑ ∑Here η is a Hermitian operator which introduces an operator measure in Hilbert space. In bothcases, the norm remains constant in time if the Hamiltonian of the system is a Hermitianoperator. If η is not a Hermitian operator, the norm still conserves the constant. Our interpre-tation: The operator η can be thought as a bivalent tensor which transforms a Hilbert spaceinto another: Fig. 30. The operator measure η interpreted physically The approach of Riesz’s representation theorem can be used for very importantconclusions about the operator measure η and its physical interpretation. This theorem estab-lishes an important connection between a Hilbert space and its (continuous) dual space: If theXXXVIII | Negative and Complex Probability in Quantum Information
  39. 39. http://vasil7penchev.wordpress.comunderlying field is that of the complex numbers as is in quantum mechanics, the two are iso-metrically anti-isomorphic. All the physical quantities can be interpreted as such operator measures η, restrict-ing them into the set of Hermitian operators. All the visible world of many different and di-vided things is that set, i.e. a clearly restricted class of the transformation of the whole intoand within itself. A natural question is about the interpretation rather of the full than restrictedclass of such transformations. The answer is: their states, i.e. all the Ψ-functions, the truewhole, are represented by Hilbert space: Moreover, it is that set which transformations formin and of itself, or mathematically speaking, are isomorphic to itself. That it is the case may bedemonstrated by the following mapping: In general, one-to-one correspondence is valid between any operator in Hilbertspace and a point in it: ( ). A tensor of any finitevalence k+l can be represented as an operator in Hilbert space and consequently, as a point in ∑it: (∏ ) . The physical homogeneity of the imaginary member, namely ( ), or cor- ( )respondingly, , needs to be interpreted, and that returns us to Einstein’s two “in-compatible” views both of discrete quanta and of differentiable, consequently continuous,mechanical motions (morphisms).The energy-Janus of two faces, combines the two: − poten-tial energy depending only on coordinates, and in that way suggesting a discrete concept ofbodies, and also kinetic energy depending solely on speeds, and in this way suggesting a con-tinuous concept of motions of those bodies. We should emphasize two innovative or even revolutionary points: 1. The wave-particle dualism of quantum mechanics already existed implicitly inclassical mechanics as both aspects of a body at rest or in motion: that is its potential or kinet-ic energy. It is the Planck constant and Heisenberg uncertainty that force the two aspects totransform between themselves (it being impossible to give separated distinctions), and then tobe added as in classical mechanics and physics. The case is such that it points to a hiddencommon essence behind the macro visible, i.e. a body and its motion. It is as if quantum mechanics for the Planck constant is concentrated on theboundary between them, which is demonstrated to be rather an area than a contour, withinwhich particulate body and continuous wave motion are the same, namely a single quantum Negative and Complex Probability in Quantum Information | XXXIX
  40. 40. Vasil Penchevobject. Negative probability describes the immediate interaction of probabilities, which makestheir relation that which is necessary information to be a physical entity. 2. The above, pinpointed, physical homogeneity of the imaginary member for thephysical quantity of energy hints that not only mass and energy, but also time share a commonessence, and consequently their mutual conversion, in the same way as that mutual conversionof the formerly two, might or even should be admitted. In the spirit of a Skolemian relativity (Skolem 1970: 138; Пенчев 2009: 307-325),any entangled subsystem of a system can be represented equivalently both as an independent,isolated, indivisible system and as an arbitrary operator (an operator measure η) transforminga point of the Hilbert space of the one system into that of another system. The conclusion made above implies particularly (entanglement = 0) that anymeasured value of a quantity in the system (‘an apparatus – a quantum object’) is ‘the objec-tive value’ of that quantity of the quantum object. There are no hidden parameters which dic-tate deterministically among its random values in this case. The Hermitian character of any phys-ical quantity generalizes the requirement that the value is in an exactly given point of time. Italso generalizes the requirements about discrete functions. Ψ-function represents the condition that a quantum quantity can obtain non-zerovalues only on areas whose common measure is zero. The absence of hidden parameter is dueto the zero measure of any area with non-zero values. The same case is represented by Dirac’sδ-function. The zero measure of any area with non-zero values is a mathematical way to rep-resent uncertainty relations. Any Ψ-function can be interpreted as the operator measure η of aquantum object entangled with its environment. V. A few mathematical questions Let us consider:  Bartlett’s approach for introducing negative probability by means of the char-acteristic function of random quantities;  Gleason’s theorem about the existence of measure in Hilbert space;  Kochen – Specker’s theorem again, but now in a statistical interpretation. Bartlett’s properly mathematical approach to negative probability is the following: “Since a negative probability implies automatically a complementary probabilitygreater than unity, we shall reconsider with all re-strictions on the values of the individual pr removed, provided that the sum remains finiteequal to the conventional sum of unity. For those familiar with the correspondence between XL | Negative and Complex Probability in Quantum Information
  41. 41. http://vasil7penchev.wordpress.comprobability theory and the theory of measure, it is noted that the parallel extension in thismore general form of probability theory corresponds to the use of an additive set functionwhich is always the algebraic difference of two positive functions” (Bartlett 1944: 71-72). Further, the interval [0, 1] of probabilities conserves a unique meaning: “Thus probabilities in the original range 0 to 1, as we might reasonably expect,still retain their special significance. It is only these probabilities which we can immediatelyrelate with actual frequencies; it is only these probabilities, for example, for which the theo-retical frequency ratio r/n tends to p with probability one, as n tends to infinity” (Bartlett1944: 72). Consequently, the rest, or nonstandard, probabilities cannot be defined within theusual understanding of whole and part: Their introduction implies necessarily “external” parts(examined above) of a whole like a quantum system. Speaking purely mathematically andreturning back to Bartlett’s proper approach, we are going to introduce such kinds of general-ized probabilities indirectly, namely as corresponding to an appropriate generalization of ran-dom quantity in such way that a random quantity for any characteristic function is required: “Random variables are correspondingly generalized to include extraordinary ran-dom variables; these have been defined in general, however, only through their characteristicfunctions” (Bartlett 1944: 73). As Dirac, Bartlett also suggested negative probability only “in the balance” ofprobabilities. Implicitly, empirical and physical reality can be only the usual type of wholeconsisting only of normal (internal) parts, with respect to which, probabilities within [0, 1]could be actually observed in experiments: “Negative probabilities must always be combined with positive ones to give anordinary probability before a physical interpretation is admissible. This suggests that wherenegative probabilities have appeared spontaneously in quantum theory it is due to the mathe-matical segregation of systems or states which physically only exist in combination” (Bartlett1944: 73). However why not make the Ψ-function be the characteristic function of a randomquantity? Bartlett’s approach leads to the idea of considering the Ψ-function as the character-istic function of a physical quantity which is “random”, or more exactly, a function of thecoordinates in configuration space. The utilization of the Ψ-function as the characteristicfunction instead of the probabilistic distribution of random quantity has the advantage of de-scribing its behavior in general: including also a discrete change of probability (a quantumleap) when the probabilistic distribution itself in that point is represented by δ-function. In Negative and Complex Probability in Quantum Information | XLI
  42. 42. Vasil Penchevfact, the discrete change of probability is what is available in all the phenomena of entangle-ment: when the probabilistic distribution of a quantum object restricts immediately the de-grees of freedom of another; as a result of that “informational” interaction, the probability of agiven point endures a discrete leap in the general case. The differential value in that point ofthe probabilistic distribution is ∞, and the most significant condition is that it cannot have afinite norm, conventionally accepted as unity. It should be > 1! If however the differentialprobability is > 1 in the point of the discrete change of probability, then it implies the appear-ance and introduction of negative probability following Bartlett’s approach: p=p1+ p2. If p1 >1, keeping p=1, then p2 < 0! Consequently in the final analysis, the appearance of negativeprobability is due to the availability of discrete leaps of probability in some points. It is whatstrongly suggests that just the ψ-function (it is the characteristic function) be utilized insteadof the probabilistic distribution itself of the random quantity. The question remains “How does the ψ-function comply with the characteristicfunction?” Let us see: Fig. 31. Ψ-function as the characteristic function of a random quantity: The Ψ-function (on the right) is obtained by the probabilistic distribution (on the left) as the integral is substituted by an infinite sum of constants (“trape- zoids”) necessarily → 0 for the factor 1/n, n→∞ . Our newly acquired demonstration of the Ψ-function as characteristic function ofrandom configuration-space coordinates leads us to a new perspective on Gleason’s theoremabout measures in Hilbert space: Till now we have discussed negative or complex probability as a relation ofmeasures (any of which may be a nonnegative real number). Gleason’s theorem states thatany measure of such an ordinary type in a Hilbert space of dimension more than twonecessarily conserves the orthogonality of the dimensions. The theorem inclines us to a sub-verse, logically negated idea: to generalize the notion of measure to complex (particularly,negative), and to investigate violating the orthogonality of the dimensions in Hilbert space. In XLII | Negative and Complex Probability in Quantum Information
  43. 43. http://vasil7penchev.wordpress.comother words, whether to be the “not complex” measure is adequate on any “curved” (i.e. hav-ing a non-orthogonal basis) Hilbert space of dimension more than 2? What is the significance of the exception of Gleason’s theorem, in the case of 2dimensions? The idea is that the exception, 2 dimensions, guarantees the “backdoor”, throughwhich a Skolemian type of relativity between “flat” and “curved” Hilbert space can pass, alsothe Skolemian relativity of a measure on “flat” and a measure on “curved” Hilbert space. Gleason’s theorem itself states: “Let μ be a measure on the closed subspaces of a separable (real or complex)Hilbert space ℋ of dimension at least three. There exists a positive semi-definite self-adjointoperator T of the trace class such that for all closed subspaces A of ℋ μ(A) = trace (TPA),where PA is the orthogonal projection of ℋ onto A” (Gleason 1957: 892-893). Bell’s interpretation (1966) of Gleason’s theorem (1957) was the following: “… ifthe dimensionality of the state space is greater than two, the additivity requirement for expec-tation values of commuting operators cannot be met by dispersion free states” (Bell 1966:450). In other words, Bell (1966) interpreted Gleason’s theorem according to the yet to cometheorem of Kochen and Specker (1967). “It was tacitly assumed that measurement of an observable must yield the samevalue independently of what other measurements may be made simultaneously” (Bell 1966:451). Otherwise, Bell’s objection (1966) to Gleason’s theorem can cover the theorem ofKochen and Specker (1967). Now we have a necessary justification to look in a new way at Kochen – Speck-er’s theorem: “statistically” interpreting it. In fact, the theorem has a clearly expressed “anti-statistical” meaning: quantum mechanics uses probabilistic distributions, which are not statis-tics. According to the discussion so far, there should exist states of negative probability, andthat’s why they cannot be “states” in a restricted, properly statistical sense. They representimmediate interactions of statistical states, i.e. the distributions of system states. The system isnot a priori or independent of its states. The parts of a system are not the identifiable with “thesubstance” of its states. Instead of conclusions (as the necessary conclusions have mentioned in the ab-stract and the beginning of the paper), a few questions: 1. Whether negative probability is only a mathematical construction, or there existphysical objects of negative probability? 2. Whether negative probability and pure relation (such a one which cannot be re-duced to predications) are equivalent, expressing the same case in different ways? 3. Whether negative probability does imply the physical existence of probability? Negative and Complex Probability in Quantum Information | XLIII
  44. 44. Vasil Penchev 4. Can probabilities interact immediately (i.e. without any physical interaction ofthe things, phenomena, or events possessing those probabilities)? 5. Whether physical, existing information is equivalent to the interaction of prob-abilities?LITERATURE:Bartlett, M. 1945. Negative Probability. – Mathematical Proceedings of the CambridgePhilosophical Society. Vol. 41, No 1 (June 1945), 71-73 (cited or mentioned on the following pagesabove: XXXIX-XLI).Bell, J. 1964. On the Einstein ‒ Podolsky ‒ Rosen paradox. ‒ Physics (New York), 1, 195-200. Alsoin: Bell, J. Speakable and unspeakable in quantum mechanics: collected papers in quantum mechan-ics. Cambridge: University Press, 1987, 14-21.( cited ormentioned on the following pages above: I, XI, XXI, XXIV, XXV, XXVIII, XXXII).Bell, J. 1966. On the Problem of Hidden Variables in Quantum Mechanics. ‒ Reviews of ModernPhysics. Vol. 38, No 3 (July), 447-452. Also in: Bell, J. Speakable and unspeakable in quantummechanics: collected papers in quantum mechanics. Cambridge: University Press, 1987,1-13. ( ; cited or mentioned on the following pages above: XI, XX, XXI, XLII.)Cattaneo, G., M. Chiara, R. Giuntini, F. Paoli. 2009. Quantum Logic and Nonclassical Logics. – In:Handbook of Quantum Logic and Quantum Structures. Quantum Logic (eds. K. Engesser, D. Gabbay,D. Lenmann). Amsterdam, ets.: Elsevier, 127-226 (cited or mentioned on the following pages above:XI).Cohen, L. 1966. Generalized phase-space distribution functions. ‒ Journal of Mathematical. Physics.Vol. 7, No 5, 781-786 (cited or mentioned on the following pages above: VI, IX, XI-XII, XIII, XIV).Cohen, L. 1989. Time-Frequency Distributions – A Review. – Proceedings of the IEEE. Vol. 77,No. 7, 941–981, 1989. ( ; cited or mentioned on the following pages above: VI, IX, XI-XII, XIII, XIV).Cohen, L. 1995. Time-Frequency Analysis. New York: Prentice-Hall (cited or mentioned on the fol-lowing pages above: VI, VII, IX, XI-XII, XIII, XIV).Cohen, L. 2008. The Weyl transformation and its generalization. – Rendiconti del Seminario Matema-tico. Università Politecnico di Torino, Vol. 66, No 4, 259-270. – (cited or mentioned on the following pages above: XIV).Dirac, P. 1942. Bakerian Lecture. The Physical Interpretation of Quantum Mechanics. –Proceedingsof the Royal Society of London, A, vol. 180 (980; March 18 1942), 1-40. ( (cited or mentioned on the following pag-es above: XXV, XXXVI, XL.) XLIV | Negative and Complex Probability in Quantum Information
  45. 45. http://vasil7penchev.wordpress.comEinstein, A. 1905. Ist die Trägheit eines Körpers von seinem Energieinhalt abhängig? – Annalen derPhysik. Bd. 18, No 13, 639–641. – (cited or mentioned on the following pages above: XXIX,XXX, XXXVIII).Einstein, A. 1905. Über einen die Erzeugung und Verwandlung des Lichtes betreffendenheuristischen Gesichtspunkt. – Annalen der Physik. Bd. 17, No 6, 132–148. ( (cited or mentioned on thefollowing pages above: XXIX-XXX, XXXVIII.)Einstein, A. 1918. Prinziplelles zur allgemeinen Relativitätstheorie. – Annalen der Physik. Bd. 55, №4, 241-244. ( ; cited or mentioned on the following pages above: VIII, XIII, XXVII, XXX, XXXI-XXXII.)Einstein, A., B. Podolsky and N. Rosen. 1935. Can Quantum-Mechanical Description of PhysicalReality Be Considered Complete? ‒ Physical Review, 1935, 47, 777-780.( .) In Russian: А. Эйнштейн, Б. Подолски, Н. Розен.1936. Можно ли считать, что квантово-механическое описание физической реальности являетсяполным? – Успехи физических наук. T. XVI, № 4, 440-446 – . (Cited or mentioned on the following pages above: XVIII, XXI, XXV).Feferman, S. 1991. Reflecting and Incompleteness. – The Journal of Symbolic Logic. Vol. 56, No 1(March 1991), 1-49 (cited or mentioned on the following pages above: X).Feynman, R. 1991. Negative Probability. – In: Quantum Implications: Essays in Honour of DavidBohm (eds. B.Hiley, D. Bohm, D. Peat). London – New York: Routledge, 235-248( (Cited or mentioned on the following pages above: XXXVI-XXXVII).Gleason, A. 1957. Measures on the Closed Subspaces of a Hilbert Space. – Journal of Mathematicsand Mechanics. Vol. 6, No 6, 885-893. ( ; cited or mentioned on the following pages above: XXXIX, XLI, XLII.)Groenewold, H. 1946. On the Principles of Elementary Quantum Mechanics. – Physica. Vol. XII, No7 (Oct 1946), 405-460; cited or mentioned on the following pages above: XIV).Hermann, G. 1935. The circularity in von Neumanns proof. (Translation by Michiel Seevinck of"Der Zirkel in NEUMANNs Beweis", section 7 from the essay by Grete Hermann, DieNaturphilosophischen Grundlagen de Quantenmechanik. Abhandlungen der friesschen Schule, 6,1935 ‒ (cited or mentioned on the following pagesabove: XX).Kochen, S., E. Specker. 1967. The problem of hidden variables in quantum mechanics. – PhysicalReview A. Vol. 17, № 1, 59-87. ( (cited or mentioned on the following pages above:XXVIII, XXIX, XXXIII, XXXIX, XLII).Kripke, S. 1975. Outline of a Theory of Truth. – The Journal of Philosophy, Vol. 72, No. 19,Seventy-Second Annual Meeting American Philosophical Association, Eastern Division. (Nov. 6,1975), 690-716. ( ; citedor mentioned on the following pages above: IX-X). Negative and Complex Probability in Quantum Information | XLV